The economic burden of obesity

October 2010

NOO | DATA SOURCES: KNOWLEDGE OF AND ATTITUDES TO HEALTHY EATING AND PHYSICAL ACTIVITY 1
NOO is delivered by Solutions for Public Health
Executive summary
Estimates of the direct NHS costs of treating overweight and obesity, and related
morbidity in England have rangeda from 479.3 million in 19981 to 4.2 billion in 2007.2
Estimates of the indirect costs (those costs arising from the impact of obesity on the
wider economy such as loss of productivity) from these studies ranged between 2.6
billion1 and 15.8 billion.2

Modelled projections suggest that indirect costs could be as much as 27 billion by

2015.2 In 2006/07, obesity and obesity-related illness was estimated to have cost 148
million in inpatient stays in England.3 In Scotland, the total societal cost of obesity and
overweight in 2007/08 was estimated to be between 600 million and 1.4 billion,4 the
NHS cost may have contributed as much as 312 million.

Whilst these figures suggest an overall increase in the costs of treating overweight and
obesity, in the absence of an agreed definition of costs, different studies have scoped
and defined costs differently. It is therefore difficult to interpret trends and to
compare cost estimates between studies.

Reports by the National Audit Office (NAO),1 the House of Commons Health Committee
(HCHC)5 and Foresight2 still underpin the majority of publications which have been
published about the NHS and wider cost of obesity in the UK.

Introduction
The increasing prevalence of obesity amongst adults and children is a major public
health challenge both nationally and internationally. Being overweight or obese can
increase the risk of developing a range of other health problems such as coronary heart
disease (CHD), type 2 diabetes, some cancers, stroke and reduce life expectancy. The
consequences of obesity are not limited to the direct impact on health. Overweight and
obesity also have adverse social consequences through discrimination, social exclusion
and loss of or lower earnings, and adverse consequences on the wider economy
through, for example, working days lost and increased benefit payments.

Background
The NHS costs of overweight and obesity had previously been estimated at 991 million
to 1,124 million5 in 2002 and the total impact on employment as much as 10 billion
for the same time period. The 2007 report from Foresight (Tackling Obesities: Future
Choices Project Report) provided an overview of obesity in the UK which included
modelled estimates of future trends in levels of obesity and obesity-related diseases,
and associated costs in terms of both the health service and of wider society.6

The purpose of this briefing note is to:

z Summarise key baseline documents
z Highlight and summarise any documents which have been published in
the last three years which may provide more up-to-date figures

a. The figures presented here come from different reports and are not directly comparable as they use
different methods and include different costs.

NOO | The economic burden of obesity 1

 z Provide options for updating estimates for the economic cost of obesity

Methods
Literature searches of relevant medical databases (Medline and HMIC), and a search of
grey literature was undertaken. Searches were limited to English language publications
from 2006 onwards. The date limit was introduced on the basis that publications prior
to 2006 are likely to have been considered as part of either the original Foresight
report literature review or in the development of the modelling methodology.

Summary of key baseline publications

Key publications and their findings are described below. A tabulated summary of the
studies, their methodology and costs is provided in the Appendix.
z Tackling Obesity in England (2001) 1 This National Audit Office (NAO)
report provided an overview on the causes, prevalence, costs and the
management of obesity in England. Data on the economic burden of
obesity was commissioned from the Dept of Economics, City University.
The direct (healthcare) costs of obesity and obesity related diseases were
derived from population attributable fractions (PAF)b obtained from a
literature review which were then applied to GP consultation rates,
hospital admissions and outpatient attendances. The indirect costs were
restricted to earnings lost due to premature mortality and lost earnings
due to obesity attributable sickness. Years of working life lost were
estimated by applying data on obesity attributable deaths (from the
literature review) to age- and sex-specific death rates and residual life
expectancy in working aged adults. Mean annual earnings data was
then applied to working life lost to estimate lost earnings. Lost earnings
due to obesity attributable sickness was derived by applying mean daily
earning figures to obesity and obesity-related work absences calculated
from information on certified incapacity benefit provided by the then
Dept of Social Security.
The cost of treating obesity in England in 1998 was estimated at 9.4
million; the cost of treating disease attributable to obesity was
estimated at 469.9 million. Both were likely to be an underestimate.
The cost of lost earnings due to premature mortality was estimated at
827 million; the costs of obesity and obesity-related lost earnings were
estimated at 1,322 million. A total projected figure of 3.6 billion was
given for 2010.
z House of Commons Health Committee (HCHC) Third Report of
Session 03/04 (2004)5 This report applied the same general principles
and methodology as the NAO report, updated to use 2002 data sources
and an extended list of obesity- related diseases. The estimated costs of
obesity in 2002 compared to 1998 are summarised in Table 1.

b. The PAF is the proportion of cases/disease attributable to a particular exposure or risk factor in a given
population.

NOO | The economic burden of obesity 2

 Table 1: Estimated direct and indirect costs of obesity (1998 and 2002)

Estimated costs 1998 ( millions) 2002 ( millions)

Treating obesity 9.4 45.8 49.0c

Treating consequences of
469.9 945 1,075d
obesity

Total direct costs 479.3 990.8 1,124

Lost earnings due to

827.8 1,050 1,150
premature mortality

Lost earnings due to

1,321.7 1,300 1,450
attributable sickness

Total indirect costs 2,149.5 2,350 2,600

Total economic cost of

2,628.9 3,340 3,724
obesity

Using the crude assumption that the costs of being overweight are on
average only half of those of being obese then, based on the estimate
that the prevalence of overweight is twice that of obesity, the cost
would double. The total economic cost of for overweight and obesity
would therefore be 6.6 to 7.4 billion per year.
z Tackling Obesities: Future Choices Project Report (2007).2 The
estimates used in the Tackling Obesities: Future Choices report was
based on a micro simulation model.2 In simple terms, this model derived
estimated projections of BMI distribution and obesity-related diseases
based on an initial analysis of Health Survey for England (HSE) annual
datasets from 1993 to 2004 and population projections. Using a disease-
cost model, the implications for NHS expenditure were estimated and
projected for 2007, 2015, 2025 and 2050. Using this model and
assumptions regarding the relationship between NHS and total costs
from the HCHC report it was estimated that in 2050, the NHS cost
attributable to obesity and overweight would be 9.7 billion and the
total costs would be 49.9 billion at 2007 prices.

What has been published since 2007 which may contribute to the
evidence?
The majority of publications relating to the cost of obesity have been generated from
North America with a handful from Europe. Since the publication of the Future Choices
report there seems to have been relatively little readily accessible data which relates to
the future NHS costs of obesity and overweight. Publications which do relate to this
specific area frequently refer back to, or are based on either the Foresight report or the
HCHC and NAO reports which contributed to the modelling. One such report Healthy

c. The 4 fold increase in the cost of treating obesity was attributable largely to increased drug costs.
d. 390 million 435 million of the increase is due to the inclusion of new obesity related diseases.

NOO | The economic burden of obesity 3

Weight, Healthy Lives: a toolkit for developing local strategies7 was published in 2008.
This report includes estimates at individual primary care trust level of the annual cost of
diseases related to overweight and obesity. The cost estimates are based on those in
the Foresight report and relate to the years 2007, 2010 and 2015.

Other publications which may consolidate and/or add to current knowledge on the
economic burden of obesity are described below.

A review8 published shortly before the Foresight report reviewed cost studies of
overweight and obesity in the UK. These were limited to the NAO and HCHC reports
and an earlier study9 which estimated the excess costs attributable to overweight and
obesity in the north west of England. The authors also developed their own method to
estimate NHS costs. This involved the application of PAF data from the WHO Burden of
Disease Project to NHS costs for overweight and obesity-related disease based on
figures from a 1996 DH discussion document updated to 2002. Using this method, the
NHS costs in 2002 were estimated to be 3.23 billion, equivalent to 4.6% of total NHS
expenditure. This figure seems high relative to other estimates and may be due to the
methodology used. Also, the WHO Burden of Disease group provide PAFs for obesity
on a country groupings basis (e.g. Northern Europe), so the figures used are not specific
to the UK.

Dr Foster Research published a report in 20083 which attempted to estimate the acute
costs of obesity in England using HES for April 2006 March 2007. It is unclear whether
a population attributable fraction was applied to the numbers of hospital stays (spells)
or whether the stays were based on identifying records with a dual diagnosis code of
an obesity-related disease plus obesity. The latter method, in particular, is likely to
result in an underestimate due to coding deficiencies. However, on the basis of the
data extracted and the application of 2007/08 cost data (source unknown), the report
estimated that during 2006/07 there were 68,627 obesity-related diagnosis stays with
an NHS cost of nearly 148 million.

The Scottish Government has published a strategy for the prevention of obesity4 which
provides direction for national and local government decision making in the short and
medium term. The report provided the following key figures on the cost of obesity in
Scotland in 2007/8:
z More than 175 million of the cost of obesity was direct NHS costs
(equivalent to 2% of the budget allocated to Health Boards) of which
4.5 million were drug costs. Nearly half the cost was attributable to
obesity-related type 2 diabetes (48 million) and hypertension (38
million).
z Using the Foresight report assumptions, the NHS cost of being
overweight and obese could be in the region of 312 million.
z Lost earnings due to premature mortality were estimated to be 87
million.
z Lost earnings due to obesity and obesity-related illness were estimated
to be 195 million. Total indirect costs were therefore estimated at 282
million.
z Estimates of the total cost of obesity to society in 07/08 could be 0.6
billion to 1.4 billion.

NOO | The economic burden of obesity 4

The costs were calculated using the same principles as the NAO and HCHC reports i.e.
based on healthcare activity data, sickness absence and estimates of premature
mortality and future obesity prevalence.10

A recent systematic review11 which aimed to assess the current published literature on
the direct costs associated with obesity concluded that obesity was estimated to
account for 0.7% (from a French study) to 2.8% (from a US study) of a countrys total
healthcare expenditure. A third study estimated that including overweight, the
proportion of expenditure was in the region of 9.1% (Canada). The modelling
undertaken to inform the Foresight report quoted a figure of 6% as the projected
proportion of NHS cost for 2007. There were no studies from the UK included in the
review; most were based on the US healthcare system; five papers were based on data
from European countries. The studies used both database and modelling methods.

The recently published report by the National Heart Forum (NHF) on adult obesity
trends12 in England updates (based on a further three years HSE data) and develops
further the micro simulation model used to predict future trends in obesity and obesity-
related disease. In contrast to future projections of obesity trends in children, obesity
trends in adults were predicted to be only slightly less than those predicted in the
Foresight report. The resource implications were not calculated.

Further considerations
NICE guidelines were published in 200613 which covered the prevention identification,
assessment and management of overweight and obesity in adults and children. As part
of the costing template developed to support PCTs in implementing the guideline, NICE
identified three key priorities for implementation that would have a significant
resource implication and for which costs could be estimated: bariatric surgery,
childrens services for the overweight and obese, and training. The national costs of
fully implementing the guideline were estimated to be 63.3 million in the first year,
with potential costs in year ten of 35.5 million and identified savings of 55.6 million.

The NHS costs are not, however, restricted to expenditure on activity. A survey of 150
hospital trusts in England in 200814 suggested that each Trust spent on average 60k on
specialist equipment (for example beds, chairs, hoists, operating tables and radiological
equipment with a larger weight capacity), a figure that had doubled in three years.
This equated to 10 million per annum if all the Trusts in England and Wales were
spending the same.

The NHS costs are wider than those attributable to primary and secondary healthcare
activity; there are also equipment and infrastructure costs. Organisations other than
the NHS also have to plan for and accommodate the wider costs of obesity and obesity-
related disease for example, the increasing cost of social care.15

Options for updating figures on the economic burden of obesity

Options for updating the cost of obesity include the following:
a. Use the methodological principles used by the NAO/HCHC updated to
encompass new healthcare activity, social service and wider economic cost
data, prevalence and assessments of population attributable risk.

NOO | The economic burden of obesity 5

 b. Apply healthcare cost data to the updated modelling undertaken by the
NHF and apply the same principles used in the Foresight report (updated to
reflect current literature estimates of the relationship between direct and
indirect costs) to estimate the wider economic costs.
c. Explore the direct healthcare costs of obesity in specific areas. For instance,
the excess costs of obesity-related caesarean sections could be calculated by
applying PAF and tariff costs to caesarean section figures obtained from
HES.

The methodological approach used by the NAO/HCHC may be the most meaningful
option for providing figures on the current economic burden of obesity. The estimates
are based on healthcare activity data directly attributable to obesity along with an
estimate of obesity-related costs, also derived from current healthcare activity. The
indirect costs are based on actual benefit data; literature estimates of the relationship
between direct and indirect costs provide additional figures.

The NHS costs in the Foresight report are generated from modelled estimates of the
distribution of elevated BMI in the population, based on HSE data. The indirect costs
were estimated using an assumption about the ratio of the NHS costs of
obesity/obesity-related disease to indirect costs rather than any specific modelling.

Whilst modelling is helpful, it necessarily relies on existing patterns of treatment and

assumptions about continued patterns of eating and physical activity as well as
behavioural and social responses to obesity. Dynamic modelling is seldom used in such
estimates but might be helpful here. It might also be helpful to look at alternative
scenarios as part of modelling estimates such as: obesity trends continue; obesity
continues to rise by a specified percentage per year; obesity is reduced by a specified
percentage per year.

NOO | The economic burden of obesity 6

References

1. National Audit Office. Tackling Obesity in England. London: The Stationery

Office, 2001.
2. Butland B, Jebb S, Kopelman P, et al. Tackling obesities: future choices project
report (2nd Ed). London: Foresight Programme of the Government Office for
Science, 2007. www.bis.gov.uk/assets/bispartners/foresight/docs/obesity/17.pdf
(Accessed 4 November 2010).
3. Dr Foster Research. Weighing up the burden of obesity: a review by Dr Foster
Research. London: Dr Foster Research, 2008.
www.drfosterintelligence.co.uk/newsPublications/loclDocuments/WeightingUpThe
BurdenOfObesityReport.pdf (Accessed 17 May 2010).
4. The Scottish Government. Preventing Overweight and Obesity in Scotland: A
Route Map Towards Healthy Weight. Edinburgh: The Scottish Government, 2010.
www.scotland.gov.uk/Publications/2010/02/17140721/0 (Accessed 17 May 2010).
5. House of Commons Health Committee. Obesity: Third Report of Session 2003/4.
London: The Stationery Office, 2004.
6. MacPherson K, Marsh T, Brown M. Tackling Obesities: Future Choices
Modelling future trends in obesity and the impact on health. London: Foresight
Programme of the Government Office for Science, 2007.
www.foresight.gov.uk/Obesity/14.pdf (Accessed 17 May 2010).
7. Swanton K. Healthy Weight, Healthy Lives: a toolkit for developing local
strategies. London: The National Heart Forum, 2008.
www/dh.gov.uk/en/Publicationsandstatistics/Publications/DH_088968 (Accessed 2
November 2010).
8. Allender S, Rayner M. The burden of overweight and obesity-related ill health in
the UK. Obesity Reviews 2007; 8: 467473.
9. Bagust A, Roberts B, Haycox A, et al. The additional cost of obesity to the health
service and the potential for resource savings from effective interventions. Eur J
Public Health 1999; 9: 258264.
10. Twaddle S, personal communication.
11. Withrow D, Alter D. The economic burden of obesity worldwide: a systematic
review of the direct costs of obesity. Obesity Reviews 2010 Jan 27 (Epub ahead of
print).
12. National Heart Forum. Obesity Trends for Adults: Analysis from the Health
Survey for England 19932007. London: National Heart Forum, 2010.
13. National Institute for Health and Clinical Excellence. Obesity: Guidance on the
prevention, identification, assessment and management of overweight and
obesity in adults and children (CG 43). London: NICE, 2006.
14. Thornton J. Supersize NHS. London: The Times, 10 May 2008.
15. Local Government Association. 6.3 billion bill by 2015 shows Britain is fast
becoming the 'obesity capital of the world'. LGA media release 7 October 2008.
www.lga.gov.uk/lga/core/page.do?pageId=1077596 (Accessed 21 May 2010).

NOO | The economic burden of obesity 7

Appendix

Summary of relevant key baseline documents and publications from 2006

Year and country

Report Summary of outputs to which data Methods Comments
refers
NAO (2001) Direct costs of treating England 1998 Cost of illness study undertaken by Costs relate to obesity alone as opposed to
obesity 9.4 million Dept of Economics, City University. overweight and obesity. Direct costs GP
Direct costs hospital admissions, consultations may be underestimate as
Cost of treating the day cases, outpatient attendances, used 1991/92 data. If number of
consequences of prescribing data (all from 2002/03), consultations increased at the same rate as
obesity 469.9 million GP consultations (from 1991/92), the increase in prevalence of obesity then
Total direct costs DH financial data (from 2002) and cost of GP consultations would increase
479.3 million obesity prevalence data (from HSE from 6.8 million to over 9 million.
1998). Indirect costs earnings lost Excludes data on consultations with other
Earnings lost due to due to premature mortality primary care practitioners e.g. practice
premature mortality estimated from PAFs, ONS nurses, dieticians. Cost of drugs likely to be
828 million mortality statistics and mean annual much higher as Orlistat (one of the principle
Earnings lost due to earnings data; earnings lost due to drugs used) was only licensed in late 1998.
sickness absence sickness absence estimated from Data coding for HES may have been less
1,322 million. days of certified incapacity Apr 97 reliable. Relative risk was not calculable for
Mar 98, obesity attributable a number of disease areas e.g. depression,
sickness from DSS and mean daily back pain.
earning figures,
Indirect costs may overestimate loss of
earnings as the prevalence of obesity is
higher in lower socioeconomic groups which
may also have below average earnings.
Excludes wider costs such as social service
costs and productivity losses.

NOO | The economic burden of obesity 8

 Year and country
Report Summary of outputs to which data Methods Comments
refers
HCHC (2004) Direct costs of treating England 2002 As above but using updated data Costs relate to obesity alone not overweight
obesity 45.8 million sources. Also covers wider range of and obesity. Direct costs GP costs based
to 49.0 million diseases attributable to obesity than on 1991/92 figures but increased by 50% to
Cost of treating the that covered in NAO analysis. reflect increase in prevalence of obesity.
consequences of The inclusion of additional disease
obesity 945 million consequences, increased prescribing costs
to 1,075 million and the increased prevalence of obesity
contributed the greatest to the increase in
Earnings lost due to direct and indirect costs between 1998 and
premature mortality 2002. Other comments as for NAO report.
1,050 million to
1,150 million
Earnings lost due to
sickness absence
1,300 million to
1,450 million.

Foresight NHS costs attributable England 2007 NHS cost data applied to Crude estimation of wider economic costs
(2007) to obesity and obesity- projected out to 2050 microsimulation modelled estimates based on arbitrary estimate of relationship
related disease 2.3 of elevated BMI and obesity-related between direct and indirect costs applied to
billion (2007); 7.1 disease, based on projections of a modelled estimate. Assumes no increase
billion (2050). BMI distribution from HSE data in NHS treatment costs. May be issues
NHS costs attributable (19932004). Indirect costs for about assumptions made in modelling.
to elevated BMI overweight and obesity estimated by
(overweight and multiplying total NHS costs for
obesity) and their obesity alone by 7.
consequences 4.2
billion (2007); 9.7
billion (2050).
Indirect costs of
overweight and obesity
15.8 billion (2007);
49.9 billion (2050).

NOO | The economic burden of obesity 9

 Year and country
Report Summary of outputs to which data Methods Comments
refers
Allender and NHS costs attributable England 2002 Obesity and overweight attributable WHO PAFs based on BMI of > 21 so will
Rayner to overweight and PAFs obtained from WHO EUR-A overestimate costs. 2002 costs were
(2007) obesity and obesity- region figures. 1992/93 DH figures unavailable so extrapolation had to be used
related disease 3.23 for proportion of total NHS (assumed (no further details). Costs restricted to PAFs
billion. England) costs attributable to for certain disease. Also, PAFs only
different diseases extrapolated to available on country grouping basis (e.g.
calculate 2002 costs. PAFs Northern Europe) so not specific to UK.
originating from WHO report applied
to 2002 NHS cost estimates.
Swanton Estimates of annual England 2007 Cost estimates based on those As per comments made in relation to
(2008) costs of diseases projected to 2015 presented in Foresight report (2007). Foresight report (above).
related to overweight
and obesity for
individual primary care
trusts
Dr Foster 68,627 obesity-related England 2007/08 Estimate of the acute costs of Obesity only rather than overweight and
Research diagnosis spells during obesity using HES for April 2006 obesity. Unclear whether a PAF was applied
(2008) 2007/08 with an NHS March 2007 to the numbers of hospital spells or whether
cost of nearly 148 the spells were based on identifying records
million. with a dual diagnosis code of an obesity
related diagnosis plus obesity.
Withrow and Obesity estimated to Global Systematic review (1990 June No studies of UK origin included. Mixed
Alter (2010) account for 0.7% to 2009) methodology modelling and database.
2.8% of a countrys Estimates of costs as proportion of total
total healthcare healthcare expenditure based on 3 studies.
expenditure; 9.1% for Estimate of higher medical costs based on
overweight and obesity. weighted average of 8 studies.
Obese individuals
estimated to have
medical costs 30%
higher than normal
weight peers

NOO | The economic burden of obesity 10

 Year and country
Report Summary of outputs to which data Methods Comments
refers
Scottish Treating obesity in Scotland 2007/08 Same principle methodology as As for HCHC report.
Government 2007/08 12.1 million. NAO report using NHS Scotland
(2010). Information Services Division activity
Treating obesity and its data, incapacity benefit caseload
consequences 175 figures and estimates of working life
million years lost using HCHC assumptions.
Lost earnings due to
premature mortality
87 million.
Lost earnings due to
obesity and obesity
related illness
estimated at 195
million.

NOO | The economic burden of obesity 11

Reader Information

Title The economic burden of obesity

Author(s) Liz Morgan, Monica Dent

Reviewer(s) Dr Jane Wolstenholme, University of Oxford

Dr Richard Fordham, University of East Anglia

Publication date October 2010

Target audience Commissioners of weight management services. Obesity/

physical activity/ nutrition professionals in: central
government; regional government; local authorities;
Primary Care Trusts; Strategic Health Authorities. Public
Health Observatories. Obesity and related academics

Description This paper draws together estimates for the economic

burden of obesity and overweight in terms of both direct
costs to the NHS and indirect costs to wider society.

How to cite Morgan E. and Dent M. The economic burden of obesity.

Oxford: National Obesity Observatory, 2010.

Contact National Obesity Observatory

www.noo.org.uk
info@noo.org.uk

Electronic location http://www.noo.org.uk/NOO_pub/briefing_papers

Copyright National Obesity Observatory

National Obesity Observatory

NOO | The economic burden of obesity 12